It is reported that Google has made new progress in developing its own data center chip and is in mass production in the second half of next year.

Google has made progress in developing data center chips, a development that means the company could start using new chips in 2025, a key effort the company seeks to beat its rival Amazon Cloud Services (AWS), according to people familiar with the matter.

According to foreign media reports, Google's server chip design team is currently developing two ARM-based 5nm server chips. Maple is based on an existing design from chipmaker Meiman Technologies, which has just been completed and handed over to TSMC for trial production, while Cypress is an in-house design developed by a team based in Israel.

Google's custom server chips are expected to be mass-produced in the second half of 2024 and deployed in data centers as early as 2025.

News of Google's progress in developing data centre chips comes as the company competes with Microsoft in artificial intelligence.

After spending billions of dollars on artificial intelligence research and acquisitions, Google has been upstaged in the past quarter by US artificial intelligence research company OpenAI.

OpenAI's new AI chatbot tool ChatGPT has been a hit recently, amassing more than 1 million users in less than a week after its launch. Within two months of its launch, it had 100 million monthly active users, faster than TikTok or Instagram.